When many information system managers consider whole backup disaster recovery architecture of the information system, in order to sufficiently deal with various disaster situations, a remote Disaster Recovery Site is considered to be established to remotely backup data as one more copy besides local site backup.
However, because of facing explosive increase of data, it is difficult for more and more enterprises to deal with backup of huge data. Therefore, developing online backup services is a general trend. Nevertheless, being affected by complicated network environments such as cross-border, cross-country, etc., and limited network bandwidth, efficiency of remote online backup based on direct control by client servers is low, and system and network overhead of the client servers is increased.
FIG. 1 illustrates components and operations of a prior art backup system. In present applications, when a client 110 writes data into local storage 120 (arrow 111), the data will be captured and transmitted to a backup server 130 through network (arrow 112), and the backup server 130 will apply journals to corresponding data in backup storage 140 (arrow 131). This is a traditional backup system 100, and implementation method thereof can effectively reduce data transmission volume since only journal data will be transmitted over the network. However, if the system is applied to a complicated remote network system, many cases such as network jam will still be encountered, which will affect overhead for system resources of the client 110.
In this context, in order to solve this problem, the present invention proposes a remote online backup system based on journal stager fragmented transmission, which can improve efficiency of the remote online backup and alleviate burdens of the client server.
Prior art remote online backup is to directly transmit data to remote backup sites, and this backup manner will results in huge system and network resources to be consumed in the case of complicated and slow network environment. Whereas the present invention uses a manner of local-remote Journal Stager, in which backup data is firstly buffered in manner of journal at a local place which has relatively fast speed, and then, is transmitted to a remote place. Therefore, system overhead can be greatly saved.
Further more, expansibility of the prior arts is relatively poor, in which same data need to be sent to multiple points respectively when multiple points remote backup is to be implemented; not only operations are trivial but also resources are wasted. Whereas adopting the manner of the present invention, only a flag SID of one remote site needs to be added into the backup data, and the job for sending data to new added sites operation will be implemented by a Journal Stager and a transmitter, which greatly simplifies management and network resources will be more efficiently utilized.